Add feature to import annotation fields used for other datasets in the Explorer

[sal-uva]

Would save a lot of manual work.

[sal-uva]

This is relevant again now, but I'm thinking of two different options:

How to handle annotations with filtered datasets:

• Option 1: Annotations and annotation fields are only technically saved to the original dataset, but are retrieved when opening a filtered dataset
  • Pros: Allows for filtering and annotating smaller slices without losing the connection to the original dataset. E.g. I have a dataset of 1000 lines, take a 100 random sample filter, annotate this one, and go back to the 1000-line dataset, it'll also have these annotations saved.
  • Cons: Overwriting 'original' annotations through a filtered dataset can be confusing; you're basically changing one dataset through the other, which goes against the idea of traceability.
• Option 2: Annotations are always only bound to one dataset and filtered datasets copy-paste existing annotations
  • Pros: Keeps things more clear and traceable: annotations in one dataset never affects the other.
  • Cons: Annotations that made on a filtered dataset are not available in the original dataset, which may be desirable.

My thinking now is to go with Option 2, but, as this issue originally suggested, add an 'Import/merge annotations' processor that allows you to retrieve and add annotations from another dataset to the current one. Then the functionality is still there, but is more of a conscious choice.

[stijn-uva]

@sal-uva is this now done with the merging of #428?

[sal-uva]

No it is not, it's still on my to-do list! Let me know if you have any thoughts on the above.